Effect of Hyperinsulinaemia and Insulin Resistance on Endocrine, Metabolic and Fertility Outcomes in Women With Polycystic Ovary Syndrome Undergoing Ovulation Induction

Duojia Zhang; Xinming Yang; Jian Li; Jiarui Yu; Xiaoke Wu

Disclosures

Clin Endocrinol. 2019;91(3):440-448. 

In This Article

Results

Relationships Between Baseline Demographic Characteristics and Laboratory Parameters and Hyperinsulinaemia and IR

Correlation analysis between clinical characteristics and FIN and HOMA-IR. The data showed significant positive correlations between weight, waist and hip circumferences, BMI, waist-to-hip ratio (WHR), and acanthosis nigricans score and FIN and HOMA-IR, whereas there was no correlation between the hirsutism score or acne score and FIN or HOMA-IR (Table 1).

Correlation analysis between sex hormones and FIN and HOMA-IR. As shown in Table 2, this study demonstrated significant negative observational associations between LH, LH/FSH, and SHBG and FIN and HOMA-IR and significant positive associations between FAI and FIN and HOMA-IR. There were no significant correlations between E2, P, FSH, TT, or FT and FIN or HOMA-IR. However, FT was positively associated with FIN (r = 0.110, P = 0.001) after adjusting for the influence of age. There were no significant correlations between LH or LH/FSH and FIN or HOMA-IR after correction for differences in BMI.

Correlation analysis between phenotype characteristics and FIN and HOMA-IR. We further analysed the relationships between phenotype characteristics (menstrual period and polycystic ovarian morphology [PCOM]) with FIN or HOMA. We found that the levels of FIN and HOMA-IR were significantly increased with prolonged menstrual period (P = 0.003, P = 0.001, respectively; Figure 1), but we did not find any correlation between PCOM and FIN or HOMA-IR (Figure 2).

Figure 1.

Correlation analysis between menstrual period and FIN and HOMA-IR. *P = 0.003, comparison of variables of FIN between different menstrual periods, # P = 0.001, comparison of variables of HOMA-IR between different menstrual periods

Figure 2.

Correlation analysis between PCOM and FIN and HOMA-IR. *P = 0.996, comparison of variables of FIN between different PCOMs, # P = 0.962, comparison of variables of HOMA-IR between different PCOMs

Correlation analysis between metabolism profiles and FIN and HOMA-IR. The analysis of FIN with metabolic indexes showed that FIN was significantly and positively associated with fasting glucose (r = 0.365, P < 0.001), cholesterol (r = 0.150, P < 0.001), TG (r = 0.396, P < 0.001), LDL (r = 0.120, P < 0.001), and Apo B (r = 0.255, P < 0.001) and significantly and negatively associated with HDL (r = −0.178, P < 0.001). HOMA-IR was also significantly and positively associated with fasting glucose (r = 0.495, P < 0.001), cholesterol (r = 0.182, P < 0.001), TG (r = 0.379, P < 0.001), LDL (r = 0.148, P < 0.001), and Apo B (r = 0.268, P < 0.001) and significantly and negatively associated with HDL (r = −0.129, P < 0.001; Table 3). The incidence of metabolic syndrome increased significantly along with the increase in FIN and HOMA-IR (P < 0.001 for both; Table 4).

Predictive Analysis of Hyperinsulinaemia and IR on Fertility Outcomes

Participants in the PCOSAct trial were equally divided into four FIN tertiles (FIN < 6.82 μmol/L, FIN 6.82-10.65 μIU/mL, FIN 10.66-16.76 μIU/mL and FIN ≥ 16.77 μIU/mL) and into four HOMA-IR tertiles (HOMA-IR <1.42, HOMA-IR 1.42-2.31, HOMA-IR 2.32-3.83 and HOMA-IR ≥3.84). The rates of ovulation among all women, ovulation per ovulation induction cycle, conception, pregnancy and live birth were significantly different between the four FIN tertiles after adjusting for acupuncture and clomiphene citrate treatment. In addition, the rates of ovulation among all women (P = 0.017), ovulation per ovulation induction cycle (P < 0.001), conception (P = 0.041), pregnancy (P = 0.012) and live birth (P = 0.003) were all significantly reduced with the increase in FIN (Figure 3). The rates of ovulation per ovulation induction cycle, conception, pregnancy and live birth were also significantly different between the four HOMA-IR tertiles after adjusting for acupuncture and clomiphene citrate treatment. In addition, the rates of ovulation per ovulation induction cycle (P < 0.001), conception (P = 0.037), pregnancy (P = 0.020) and live birth (P = 0.007) were significantly reduced with the increase in HOMA-IR level (Figure 4). After adjusting for age, TT and FT, with the increasing trend of serum insulin level, the ovulation, cycle ovulation, conception, pregnancy and live birth rates of patients decreased significantly (Figure 5). After adjustments for age, total testosterone and free testosterone, increasing serum insulin levels and HOMA-IR were significantly associated with decreased cycle ovulation, conception, pregnancy and live birth rates (Figure 6).

Figure 3.

Predictive analysis of FIN on fertility outcomes. P 1 = 0.017, P 2 = 0.001, P 3 = 0.041, P 4 = 0.012, P 5 = 0.003, P 6 = 0.111, respectively, compared with ovulation, ovulations per cycle, conception, pregnancy, live birth and pregnancy loss between four FIN tertiles adjusted for the treatment

Figure 4.

Predictive analysis of HOMA-IR on fertility outcomes. P 1 = 0.054, P 2 = 0.001, P 3 = 0.037, P 4 = 0.020, P 5 = 0.007, P 6 = 0.109, respectively, compared with ovulation, ovulations per cycle, conception, pregnancy, live birth and pregnancy loss between four HOMA-IR tertiles adjusted for the treatment

Figure 5.

Adjusted predictive analysis of FIN on fertility outcomes

Figure 6.

Adjusted predictive analysis of HOMA-IR on fertility outcomes

Based on a large-sample survey of Chinese patients with PCOS,[17] we regard HOMA-IR > 2.69 as the diagnostic standard of IR. The present study showed that conception (OR = 1.40, 95% CI = 1.05-1.87), clinical pregnancy (OR = 1.63, 95% CI = 1.18-2.27) and live birth (OR = 1.81, 95% CI = 1.28-2.55) were significantly lower in the IR group compared with the non-IR group after adjusting for treatment (Table 5). Our data did not show that miscarriage rates were related to FIN or HOMA-IR.

Comments

3090D553-9492-4563-8681-AD288FA52ACE

processing....